Chemical Engineering Science 101 (2013) 655–662

Contents lists available at ScienceDirect

Chemical Engineering Science

journal homepage: www.elsevier.com/locate/ces

Improved dispersion of cellulose microcrystals in polylactic acid (PLA)

based composites applying surface acetylation
Tapasi Mukherjee a,n, Marco Sani b, Nhol Kao a, Rahul K Gupta a,
Nurul Quazi a, Sati Bhattacharya a
a
Rheology and Materials Processing Centre, School of Civil, Environmental and Chemical Engineering, RMIT University, Melbourne, VIC-3001, Australia
b
Bio21 Molecular Science and Biotechnology Institute, The University of Melbourne, Parkville, Melbourne, VIC-3010, Australia

H I G H L I G H T S G R A P H I C A L A B S T R A C T

Microcrystalline cellulose (MCC) was

partially surface acetylated.

Surface acetylated MCC was used as
reinforcement in poly (lactic acid)
composites.

Rheological percolation was evalu-
ated to quantify dispersion.

Surface acetylation improved disper-
sion reaching its optimal value at
2.5 wt%.

art ic l e i nf o a b s t r a c t

Article history: Design of sustainable bioplastics can be achieved by preparing composites from renewable materials like
Received 20 March 2013 microcrystalline cellulose (MCC) fibre and biopolymer such as polylactic acid (PLA). The key driving
Received in revised form factor that affects their performance is the quality of dispersion of MCC in the PLA matrix. In this study,
15 July 2013
surface modification, one way to facilitate improved dispersion, is carried out by acetyl chloride. PLA
Accepted 19 July 2013
composites were prepared with the acetylated MCC applying solvent casting technique. Confirmation of
Available online 26 July 2013
acetylated group is accompanied by FTIR and NMR study. Change in crystalline property and thermal
Keywords: behaviour is observed by XRD study. Improvement in storage modulus (G′) is reflected in shear
Dispersion rheological tests, reaching an optimal value at 2.5 wt%. This improvement is primarily attributed to a
Polymers
more homogeneous dispersion of MCC in the matrix. Rheological percolation threshold is calculated to
Polymer processing
quantify the level of dispersion. This study is aimed to quantify the level of dispersion of acetylated MCC,
Rheology
Cellulose microcrystals as compared to pure MCC by shear rheology.
Poly(lactic acid) Crown Copyright & 2013 Published by Elsevier Ltd. All rights reserved.

1. Introduction
Microcrystalline cellulose (MCC) fibre, as the name suggests, is
commercially available and has gained significant interest in the
last decade (Lin et al., 2011; Frone et al., 2011). Some of the unique
n
Corresponding author. Tel.: +61 3 9925 2410.
E-mail addresses: tapashi.mukherjee@rmit.edu.au (T. Mukherjee),
msani@unimelb.edu.au (M. Sani), nhol.kao@rmit.edu.au (N. Kao),
rahul.gupta@rmit.edu.au (R. Gupta), nurul.quazi@rmit.edu.au (N. Quazi),
sati.bhattacharya@rmit.edu.au (S. Bhattacharya).
features of MCC include its morphology, low density and mechan-
ical strength (Huda et al., 2006). Polylactic acid (PLA) based
cellulose composites have been much explored in the recent years
as the key material to develop the next generation of light weight
and high performance materials for a variety of defense, infra-
structure and energy applications (Jonoobi et al., 2010; Peterson
et al., 2007; Pei et al., 2010). The necessity to improve the
dispersion of cellulose microcrystals (MCC) in biopolymer matrix
like polylactic acid (PLA) has become increasingly important to
enhance the mechanical properties of such composites and
thereby the commercial viability of such products for biomedical

0009-2509/$ - see front matter Crown Copyright & 2013 Published by Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.ces.2013.07.032
656 T. Mukherjee et al. / Chemical Engineering Science 101 (2013) 655–662
as well as flexible packaging applications (Braun and Dogan,
2009).
Dispersion is the key challenge to prepare such composites
(Braun and Dogan, 2009). The presence of hydroxyl groups on the
surface of cellulose restricts its homogeneous dispersion in PLA, by
initiating agglomeration or entanglement (Dubief et al., 1999). A
possible strategy to overcome this challenge is surface modifica-
tion, where the hydroxyl group is partially replaced by another
functional group. In general, surface functionality of cellulose
microcrystals as carried out from the literature review study, can
be broadly categorized into three groups; (I) native surface
chemistry of the particle as a result of their extraction like acid
hydrolysis by sulfuric acid (Frone et al., 2011) (II) physical adsorp-
tion of surfactants or polyelectrolytes (Oksman et al., 2006) and
(III) covalent modification such as esterification/etherification
Braun and Dogan (2009), silylation (Pei et al., 2010) and polymer
grafting (Pracella et al., 2010).
In this work the hydroxyl groups, as present on the surface of
cellulose was partially substituted by acetyl groups by using acetyl
chloride at room temperature. The primary motivation behind this
study was to improve the dispersion of MCC in PLA matrix by
surface acetylation. Emphasis was laid to characterize the behaviour
of dispersion at different level of loadings, ranging from 1 to 5 wt%.
Beyond 5 wt% loading, filler–filler interactions were observed.
Composites were prepared with this surface acetylated MCC, using
PLA as base matrix by solution casting technique and dichloro-
methane as the solvent (Mukherjee et al., 2012). Acetylation is
confirmed by FTIR and NMR study. The behaviour of dispersion is
characterized by XRD, DSC, and shear rheological tests. A rheologi-
cal percolation threshold is calculated to quantify the level of
dispersion and the optimal loading for a uniform dispersion.
2. Experimental section
2.1. Material
A polylactic acid biopolymer (Nature Work PLA Polymer
4032D) with a density of 1.24 g cm  3 and a melting point of
160 1C was chosen as matrix. A microcrystalline cellulose (MCC)
with a mean particle size of 20 mm, supplied in powder form by
Sigma Aldrich was used as a raw material. Acetic anhydride,
pyridine, dichloromethane was purchased from Sigma Aldrich
(Mukherjee et al., 2012).
2.2. Sample preparation
2.2.1. Acetylation of MCC
Acetylation was performed with constant stirring in a 100 ml
round bottom flask. A suspension of 2 g of MCC and 20 ml of
dichloromethane (DCM) was dispersed by constant stirring for half
an hour. A volume of 5 ml acetyl chloride was added to the flask
with 1 ml of pyridine in an inert nitrogen atmosphere. The reaction
was kept at room temperature for 48 h. After the reaction, the
product was isolated by precipitation in 20–50 ml ethanol and then
filtered. The final product was then washed 3–5 times with ethanol/
acetone to eliminate excess of acetyl chloride and pyridine. Finally
acetylated cellulose microcrystals coded as AC-MCC were oven
dried in vacuum at 70 1C for 24 h (Mukherjee et al., 2012).
2.2.2. Preparation of composites.
The desired amount of MCC/AC-MCC was added along with
20 g of PLA into DCM solvent (200–400 ml approx) with mechan-
ical stirring to produce a mixture. It formed a viscous solution over
a period of 5 h. The mixture was then conditioned overnight to
eliminate bubbles and was then cast into a petridish. DCM was
allowed to evaporate off at ambient temperature (ca. 25 1C) for
approx. 24 h. Finally, the solidified films, with a thickness of about
0.2 mm, were vacuum dried overnight, and then kept in a
desiccator containing silica gel. The resultant composite sheets
were coded as PLA-MCC/AC-MCC-1.5, PLA-MCC/AC-MCC-2.5, PLA-
MCC/AC-MCC-3.5 and PLA-MCC/AC-MCC-5, the Arabic numerically
representing the MCC/AC-MCC content in the PLA based matrix
(Mukherjee et al., 2012).
2.3. Characterization
2.3.1. FTIR
FTIR-KBr spectroscopic studies were carried out in drift mode
on the samples with a PerkinElmer FTIR spectrophotometer
(TA 8000). A total of 32 scans per sample were taken, starting
from 4000 to 450 cm  1, with a resolution of 4 cm  1. The spec-
trum was analysed with Perkin Elmer Spectrum Software
(Mukherjee et al., 2012). The samples for FT-IR analysis were
prepared by grinding the dry blended powders with KBr, often in
the ratio of 1:100(Sample: KBr), using a pestle and a mortar. For
the drift spectra, a Perkin-Elmer diffuse reflectance (DR) sampling
cell was used.
2.3.2. 13C solid state NMR
The CP/MAS 13C-NMR spectra were recorded on a VNMRS
600 MHz NMR spectrometer equipped with 4 mm CPMAS triple
resonance probe. The 13C frequency was 150.8 MHz and the chemical
shift resonances were referenced to the adamantane chemical shifts
obtained with a similar pulse sequence. Acquisition was performed
with a CP pulse sequence using a 5.25 ms proton π/2 excitation pulse, a
spectral width of 50 kHz, an acquisition time of 25 ms using a SPINAL-
64 decoupling scheme of 70 kHz strength and a 5 s delay between
transients. The Hartmann–Hann match was optimized on the  1
sideband at a MAS speed of 8 kHz with a contact time of 500 ms and a
linearly ramped-CP field of 3070.8 kHz. Typically, 12,000–16,000
transients were accumulated at 251C. A line broadening of 50 Hz
and zero filling to 65k points were used to process the spectra
2.3.3. Morphological tests
Morphology of MCC and the fractured surface of the MCC based
PLA composites at different loading of weight percentage, were
examined using an FEI Quanta 200 environmental scanning
electron microscope (ESEM), at an acceleration voltage of 30 kV
and pressure of 0.50 Torr. MCC powder was sprayed on a carbon
tape to study its morphology. The fractured surface was prepared
by immersing the specimen in liquid nitrogen and manually
fracturing the specimen. Samples were then mounted on to a
sample holder with the fractured surface facing upwards. The
sample holder was then placed in the ESEM sample chamber.
Various sample surfaces were scanned to obtain a visual impres-
sion of fibre fracture, distribution and the appearance of the fibre/
polymer interface (Mukherjee et al., 2012).
2.3.4. Crystallinity study by X-ray diffraction (XRD)
XRD patterns were obtained using a Bruker D4 Endeavor X-ray
diffractometer in the angular range of 6–901 (2θ) at a voltage of
40 kV and current of 100 mA. Peak intensities were every 0.021 at
sweep rates of 1.01 2θ/min, to see the change in crystallinity after
acetylation if any.
2.3.5. Differential scanning calorimetry (DSC)
Thermal analysis was performed on a DSC-2920 Modulated DSC
(TA Instruments) calibrated with indium with a nitrogen atmo-
sphere, with a flow rate of 35 ml/min. An empty aluminium pan
served as a reference. Composite sample weights of approximately
 T. Mukherjee et al. / Chemical Engineering Science 101 (2013) 655–662
8–10 mg were employed and the samples were encapsulated in
aluminium pans. The samples were heated and cooled at a rate of
20 1C/min in the first heating–cooling cycle and then at 2.00 1C/min
in the second heating–cooling cycle. In the first cycle, samples were
cooled to  20 1C, followed by heating to 220 1C. The samples were
held at  20 and 220 1C for 5 mins in each cycle. Glass transition
temperature (Tg), melting temperature (Tm), and the heat of fusion
(ΔHm) were taken from the second heating curve. The first heating–
cooling curve was used to erase prior thermal history in the
specimen. The cold crystallization temperature (Tcc) was considered
as the minimum of the exothermic peak. The area under the curve
was calculated as the enthalpy from the instrument software. The
degree of crystallinity (Xc) of PLA-composite can be calculated based
on the enthalpy value of a 100% crystalline PLA sample from Eq. (1).
The enthalpy of fusion of 100% crystalline PLA for Natureworks 4032
D is considered as 93 (J/g) (Mathew et al., 2006).
ΔHmΔHc
ΔXc ¼  100
93ð1XmccÞ
where Xmcc is the amount of MCC/AC-MCC in the sample.
2.3.6. Shear rheology
The tests were conducted using 25 mm parallel plate at
temperature of 170 1C. All measurements were performed using
a force transducer with a range of 0.2–200 g-cm torque. Prior to
any tests, the zero gap between the parallel plates was calibrated
at the required temperature. Dynamic strain sweeps were con-
ducted to check the linear viscoelastic region (LVR) of these
composites. Dynamic frequency sweep tests were conducted
between the frequency of 0.1 and 100 rad/s at 170 1C. Dynamic
strain sweep measurements were carried out at 1 and 10 rad/s at
0.1–100% strain, to measure the linear viscoelastic region. Dynamic
Fig. 1. FT-IR study of AC-MCC as compared to MCC with reaction scheme of acetylation of cellulose by acetyl chloride.
657
frequency sweep experiments were then performed in the linear
viscoelastic region of each material with frequency between
0.01 and 100 rad/s. A waiting time of 60 s after a loading or after
a pre-shear were used to eliminate the effects of loading history
(Mukherjee et al., 2012).
3. Results and discussion
3.1. Acetylation of MCC
Acetylated cellulose was characterized by Fourier transform
infrared (FT-IR) and solid-state cross-polarization magic angle
spinning carbon-13 nuclear magnetic resonance (CP/MAS
13
C-NMR) spectroscopy. The reaction scheme with acetyl chloride
with FT-IR absorbance spectra of unmodified cellulose and acety-
ð1Þ
lated cellulose resulting from the two reactions is illustrated in
Fig. 1. For the pure MCC, as reported in literature, a strong band
around 3, 434 cm  1 is observed which is attributed to the
stretching of hydroxyl groups (Das et al., 2010). In comparison,
the spectrum of acetylated cellulose shows successful evidence of
acetylation with ester bond appearing around 1756 cm  1. Another
important aspect in the acetylated MCC spectrum is the decreasing
absorption peak at 3, 434 cm  1, assigned to stretching vibration of
hydroxyl groups, as compared to pure MCC (Mukherjee et al.,
2012). Fig. 2 represents the spectral curve by 13C solid state NMR.
With acetylation, there were two chemical shifts located at 172
and 20 ppm, as assigned to –C ¼0 and –CH3 stretching, thus
confirming the successful acetylation of MCC. The results obtained
were similar to the observation made by Lin et al. (2011), where
acetylation was performed on cellulose nanocrystals.
658 T. Mukherjee et al. / Chemical Engineering Science 101 (2013) 655–662

3.2. Morphological study

Environmental scanning electron microscopy (ESEM) was used

to investigate the morphology of the cellulose microcrystals
embedded in PLA matrix. The untreated MCC in Fig. 3(a) had a
rod like morphology with an aspect ratio of 4.5–7.5. After acetyla-
tion of MCC, the rodlike shape was still preserved as shown in
Fig. 3(b). But its size slightly decreased, with a reduced aspect ratio
of 3.5–6.5. From the image, it is also evident that the image of
acetylated MCC is slightly blurry, which is possibly due to the
partial solubilization of cellulose molecules, similar to the obser-
vation made by Lin et al. (2011), for acetylated cellulose nanocrys-
tals, retaining its intrinsic morphology. Although there is no
morphological change as such as involved in acetylation, however,
the partial substitution of the hydroxyl groups with the carbonyl
groups on the surface of MCC particles is believed to likely initiate
13
Fig. 2. C CP-MAS NMR spectra of pure MCC and acetylated MCC. a repulsive force between adjacent MCC particles and reduces the

100 µm 100 µm
100 µm

MCC AC-MCC

MCC

200 µm 200 µm 200 µm

AC-MCC

MCC
AC-MCC

200 µm 200 µm
200 µm

Fig. 3. (a) Morphology of MCC, (b) morphology of AC-MCC, (c) neat fractured surface of PLA, (d) PLA–MCC (1.5 wt%), (e) PLA–AC-MCC (1.5 wt%), (f) PLA–MCC (2.5 wt%),
(g) PLA–AC-MCC (2.5 wt%), (h) PLA–MCC (5 wt%), (i) PLA–AC-MCC (5 wt%).
 T. Mukherjee et al. / Chemical Engineering Science 101 (2013) 655–662 659

cluster formation or agglomeration. The rod like stricture, as can made by Lin et al. (2009), where the original structure of the PLA
be seen from SEM images, can be visualized as a cluster of MCC matrix was cleaved to a greater extent. The peak around 44.21, 641
particles agglomerated together. So with the reduced agglomera- and 82.61 became more significant above that loading as well.
tion, there is a tendency to reduce the size and thereby the aspect Biopolymers like PLA contains both crystalline and amorphous
ratio (L/D) of the particles. phase arranged randomly. When acetylated MCC were introduced
Fig. 3(c–i) shows the fractured surface of neat PLA and to the PLA matrix, the uniform distribution of rigid microcrystals
composites reinforced with pure and acetylated MCC at different increased the crystalline property of the composites, as indicated by
loading. Similar to the micrographs as reported by Frone et al. the intensified peak located around 44.21, 641 and 82.61. These
(2011), the addition of MCC influenced the fracture mechanism of results can be compared to Lin et al. (2011), where the crystalline
PLA at cryogenic temperature. ESEM images show a smooth property of the nanocrystals exhibited the growth of two diffraction
fracture surface in the case of PLA (as shown in Fig. 3(c)), but an peaks located at about 16.4 and 22.6 of 2θ.
uneven fracture surface in the case of PLA composites, suggesting a
significant matrix deformation after the addition of MCC. The
3.4. Thermal analysis (differential scanning calorimetry)
images indicate that the treated fibres were better distributed in
the polymer matrix in comparison with composites obtained from
MDSC thermograms of PLA–MCC composite and PLA–AC-MCC
pure MCC. Fig. 3(d) and (e) represents PLA–MCC and PLA–AC-MCC
composites are shown in Figs. 5 and 6. The data for the glass
composites, both at 1.5 wt% respectively. Cellulose nucleation with
transition (Tg), change in reverse heat capacity (ΔCp), cold crystal-
agglomeration starts at 2.5 wt% of MCC, as indicated in Fig. 3(f).
lization (Tcc), melting temperature (Tm), heat of fusion (ΔHm), and
Agglomeration is more evident at a higher loading of MCC (5 wt%)
the degree of crystallinity within the PLA fraction (xc)of the
as shown in Fig. 3(h), which is arrested to an extent in PLA-AC-
composites are summarized in Tables 1 and 2 in the Supplemen-
MCC as shown in Fig. 3(i) (Mukherjee et al., 2012). This observa-
tary information, for the PLA–MCC and PLA–AC-MCC composites
tion is similar to the results as observed by Huda et al., 2006),
respectively. From the data available, it is evident that there is no
where they studied the SEM micrographs of silane treated talc
major change in Tg, but when change in reverse heat capacity
reinforced PLA/newspaper composites. Regularity in the fibre
(ΔCp) is calculated, it is observed that the value of ΔCp is highest in
dispersion is observed in the matrix with less large size aggregate
PLA–ACMCC composite at 2.5 wt% loading, thereby showing a
of cellulose. Frone et al. (2011), also evaluated the similar effect of
maximum constraint of polymer relaxation due to improved
surface treatment on MCC crystals in their dispersion in PLA
dispersion. It will also be interesting to interpret as to how the
matrix by using silane coupling agent. The ability of AC-MCC to
MCC dispersion affects the nucleation and growth geometry of the
reduce the filler–filler interactions through the reduction of sur-
crystals. This can be partially achieved by observing the values of
face energy can result in better filler dispersion (Lin et al., 2011).
Tcc and Xc data. Cold crystallization value of pure PLA is slightly
higher than those of the composite, where the depression trend is
3.3. Crystallinity study by X-ray diffraction (XRD)

In XRD analysis, neat PLA was mainly comprised of the amor- 0.2

phous polymer structure, with intermittent crystalline structure

characteristics, as shown in Fig. 1s in the Supplementary informa- PLA-Neat
tion. As reported in the literature, it exhibited four main different PLA-MCC-1.5 wt %
PLA-MCC-2.5 wt %
Heat Flow (W/g)

peaks located at about 16.41, 19.11, 22.61 and 44.21, with the 0.1 PLA-MCC-5wt %
appearance of a less intensified peak around 641 and 82.61. Four
main peaks correspond to different crystal structures (Furuhashi
and Yoshie, 2012; Chen et al., 2011). Acetylated MCC, when
0.0
introduced at a different loading (1–5 wt%), improved the crystal-
linity of neat PLA in overall, as indicated by growing intensity of the
peak located around 44.21, 641 and 82.61 in Fig. 4, particularly at a
loading above 1 wt%. X-Ray diffraction patterns did not change after
-0.1
the addition of the filler, indicating that the crystalline structure of 0 50 100 150 200 250
-50
the polymer is unaltered. This behavior is unlike the observation Exo Down Temperature (°C) Universal V4.2E TA Instruments

Fig. 5. DSC thermo grams of PLA–MCC composite (1.5–5 wt%).

0
1
120000 1.5 0.2
1.75 PLA-Neat
PLA-AC-MCC-1.5 Wt%
2.5 PLA-AC-MCC-2.5 Wt%
100000 3.5 PLA-AC-MCC-5 Wt%
5 0.1
80000
Heat Flow (W/g)
Intensity

60000 0.0

40000
-0.1
20000

0 -0.2
10 20 30 40 50 60 70 80 90 -50 0 50 100 150 200 250
2 Theta Exo Down Temperature (°C) Universal V4.2E TA Instruments

Fig. 4. PLA-AC-MCC composite X-ray diffraction pattern. Fig. 6. DSC thermo grams of PLA–AC-MCC composite (1.5–5 wt%).
660 T. Mukherjee et al. / Chemical Engineering Science 101 (2013) 655–662

followed, particularly after acetylation. Similar trend is observed 100000

by Reinsch and Kelley (1997), on woodfibre reinforced PHB
composite, where they concluded that the enhancement of the 10000

Storage Modulus (Pa)

nucleation rate is likely responsible for reduced cold crystallization
temperature and thereby an overall cold crystallization rate.
Calculation on percentage crystallinity applying Eq. (1) indicates 1000
that crystallinity of PLA–AC-MCC at 2.5 wt% loading is slightly
higher than rest of the composite (∼38%). This result is comparable 100
with Pei et al. (2010), where the percentage crystallinity values
increased slightly by using surface silylated cellulose nanocrystals PLA-Neat
10 PLA-MCC-5 Wt %
in PLA matrix. They concluded that the nucleating effect of the
PLA-AC-MCC-5 wt %
filler in the polymer matrix is enhanced if homogeneous disper-
sion of cellulose in poly(lactic acid) matrix is achieved. 1
0.1 1 10 100
Frequency (rad/s)
3.5. Shear rheology
Fig. 8. Comparison of storage modulus (G′) of acetylated MCC based PLA with that
Rheological measurements are complimentary to traditional of pure MCC and neat PLA at 5 wt%.

polymer nanocomposite analysis technique and may serve as an

analytical tool to differentiate the degree of dispersion (Zhao et al., 100000

2005). In this work, the quality of the degree of dispersion of MCC

Complex Viscosity h* (Pa-s)

in PLA matrix is evaluated using dynamic frequency sweep tests,
10000
by associating their storage modulus and complex viscosity para-
meters to their dispersion. In earlier work on polymer nanocom-
posites, it is well established that the rheological behavior is 1000
indicative of the level of interaction of nanoclay with the polymer
matrix and is intimately related to the clay type, its surface
100
treatment and affinity with the polymer matrix (Ray and PLA-AC-MCC-5 wt %
Okamoto, 2003; Dan et al., 2006). Figs. 2s and 3s (Supplementary PLA-Neat
information) compare the storage modulus (G′) of the PLA–MCC/ 10 PLA-AC-MCC-1.5 wt %
PLA–AC-MCC composites at various loadings, ranging from 1.5 to 'PLA-AC-MCC-2.5 wt %
5 wt% at 170 1C, when subjected to dynamic frequency sweep PLA-AC-MCC-3.5 wt %
1
tests. Figs. 7 and 8 compare the G′ of acetylated MCC with that of 0.1 1 10 100
pure MCC and neat PLA at 2.5 and 5 wt%, respectively. Fig. 4s Frequency ω (rad/s)
(Supplementary information) and Fig. 9 compare the complex
viscosity (η*) of pure and acetylated MCC at different loadings. In Fig. 9. Comparison of complex viscosity at different loading of PLA–AC-MCC(1.5–5 wt%).

all dynamic frequency sweep tests, the frequency is varied from

0.01 to 100 rad/s. From these figures it is evident that the
enhancement of storage modulus (G′) of PLA–MCC composites as
compared to neat PLA is significant at a comparatively lower level
of loading (1.5–2.5 wt%) than that at a higher level of loading. The
optimal loading reached at 2.5 wt%. From Figs. 7 and 8, it is also
observed that acetylation did improve the G′ of the composites as
compared to the composites from pure MCC as well as neat PLA up
to a loading of 2.5 wt%, in the lower frequency range sweep. At
higher loading above 2.5 wt%, there is an exponential drop in G′
value (Mukherjee et al., 2012). It is also evident from the curve,
Fig. 7. Comparison of storage modulus (G′) of acetylated MCC based PLA with that
of pure MCC and neat PLA at 2.5 wt%.
that at lower frequency, the composites behaved more solid like,
whereas at a higher frequency, the composites behaved more
liquid like. This is probably because at higher frequency, cellulose
is highly deformable, leading to decrease in storage modulus. On
the other hand, lower frequency leads to increase in storage
modulus, due to low deformation of cellulose. This phenomenon
could be possibly explained by analysing the microstructure of the
polymer matrix. Non acetylated cellulose microcrystals (MCC),
being hydrophilic in nature, has polar –OH groups that facilitate
intra as well as intermolecular hydrogen bonding in the PLA matrix.
However, PLA is a hydrophobic non polar thermoplastic polymer.
PLA and MCC thus have the tendency to repel each other and
therefore not compatible as compared to acetylated derivatives of
MCC. Creation of hydrophobic surface restricts agglomeration,
resulting in a better dispersion of MCC in the matrix. Therefore,
acetylated MCC (AC-MCC) is more compatible in the PLA matrix and
stay together in the polymer chain constraints, possibly bonded by a
weak Vanderwalls force. This results improvement in dispersion
and thereby an increased storage modulus up to a loading of 2.5 wt
%. At a comparatively higher loading, the storage modulus even-
tually deteriorates, as they affect the interactions between crystal-
line and amorphous domains of the polymer chain. When the
loading is reached beyond the saturation level of dispersion, this
Vanderwall′s force is weakened by sharing of energy. Further to the
analysis, slope of G′ at lower frequency, which is represented as n′ is
plotted against various loadings (1.5 to 5 wt%) of PLA–AC-MCC
composites, shown in Fig. 10.
This figure evaluates the rheological percolation threshold for a
uniform dispersion. From the figure, it is observed that around
2.5 wt% loading, a region of minimum slope exists and indicates
 T. Mukherjee et al. / Chemical Engineering Science 101 (2013) 655–662
Fig. 10. Slope (n′) for storage modulus G′ at low frequency against vs PLA–AC-MCC
composite at different loading.
that beyond that region, dispersion is affected by agglomeration.
This empirical study is indicated to quantify the degree of disper-
sion and evaluate the optimal loading for a uniform dispersion of
MCC in PLA matrix. Similar trend in result is observed in morpho-
logical study, where the micrograph shows the tendency of MCC
and AC-MCC to agglomerate beyond a loading of 2.5 wt%.
Complex viscosity and its reciprocal, complex melt fluidity, is a
significant parameter to study structure relationship in polymer
composites (Verney and Michel, 1989). Cox–Merz rule, a well-
known hypothesis states that the functional dependence of com-
plex viscosity′s magnitude, expressed as a function of frequency, is
identical to the functional dependence of the steady shear viscos-
ity expressed as a function of shear rate (Cox and Merz, 1958).
From Fig. 4s and Fig. 9, it is observed that complex viscosity (η*) is
slightly higher than pure PLA for all the composites up to a loading
of 2.5 wt%, indicating better dispersion of the filler. With increas-
ing loading beyond 2.5 wt%, the value of ηn is reduced, reaching its
minimal value at 5 wt%. It is further observed that PLA–AC-MCC
2.5 wt% exhibited reduced shear thinning behaviour, probably due
to chain scission mechanism. Reduction in complex viscosity at a
higher loading could be possibly due to a small decrease of
polymer molecular entanglement density as a function of the filler
aggregate, causing disruption in the polymer chain entanglement
network (Hatzikiriakos et al., 2005).
3.6. Discussions on dispersion
In this work, attempts have been made to quantify the degree of
dispersion of acetylated MCC in the PLA by applying shear rheology
and morphological study. The behavior of dispersion is further
characterized by crystallinity (XRD) and thermal study (DSC). From
the morphological study, ESEM images indicate that treated fibres
were better dispersed in PLA matrix. PLA–AC-MCC (2.5 wt%) showed
a tendency for cell nucleation with an improved dispersion. At a
comparatively higher loading (5 wt%), agglomeration was predomi-
nant, which was arrested to an extent by acetylation. These observa-
tions were complimented by shear rheological tests, where the
enhancement of storage modulus (G′) of PLA–AC-MCC composites
as compared to PLA–MCC composites is much higher at lower
frequency, particularly at a loading of 2.5 wt%. At a comparatively
higher loading, because of poor dispersion, the storage modulus
eventually deteriorates, as they affect the interactions between
crystalline and amorphous domains of the polymer chain. Analysis
on the rheological percolation threshold, points to a region of
minimum slope around 2.5 wt%, indicating that beyond that region,
dispersion is affected by agglomeration. Similar trend is observed in
complex viscosity data.
In crystallinity study with XRD, when acetylated MCC were
introduced to PLA matrix up to a loading of 5 wt%, the uniform
distribution of rigid crystals in the matrix inhibited the motion of
661
the polymer chains, and resulted in enhancement of crystalline
property of the composite, as indicated by the intensified peak
located at 16.41 and 22.61. Thermal study with DSC also leads to
the fact that PLA–AC-MCC composites have higher crystallinity,
reaching its optimal value at 2.5 wt%, possibly because of the
homogeneous dispersion of cellulose in PLA matrix.
4. Conclusions
Successful acetylation of microcrystalline cellulose (MCC) was
accompanied by acetyl chloride at room temperature as revealed
in NMR and FTIR studies. Improvement in dispersion was observed
in morphological and rheological tests, when composites were
prepared by this surface acetylation technique, particularly at a
lower loading reaching its optimum value around 2.5 wt% from the
rheological percolation threshold analysis, indicating that beyond
this region, dispersion is affected by agglomeration of the filler.
Improvement in crystalline structure was observed in XRD study,
when MCC was incorporated into the matrix. Acetylation did not
affect this crystalline structure. Improved thermal behavior was
observed in DSC thermo gram. Further analysis reveals that
percent crystallinity is enhanced for all PLA–AC-MCC composites,
in comparison to PLA–MCC a composite, reaching its optimal value
at 2.5 wt%.This suggests that AC-MCC is a stronger candidate to act
as a nucleating agent in comparison to pure MCC in preparing such
PLA based composites.
Acknowledgments
The authors would like to acknowledge the facilities, and the
scientific and technical assistance of Prof F. Seperovic for providing
us the facility to conduct solid state NMR at Bio21 Molecular
Science and Biotechnology Institute, The University of Melbourne,
Mr Phil Francis and Mr Peter Rummel, of the Australian Micro-
scopy and Microanalysis Research Facility at the RMIT Microscopy
and Microanalysis Facility, RMIT University, Frank Antolosic Mike
Allan and Muthu Panniselvam, from the Department of Chemical
Engineering and Chemistry, RMIT University, for their continued
support in completing the experimental work.
Appendix A. Supporting information
Supplementary data associated with this article can be found in
the online version at http://dx.doi.org/10.1016/j.ces.2013.07.032.
References
Lin, N., Chen, G., Huang, J., Dufrense, A., Chang, P.R., et al., 2009. Effects of polymer-
grafted natural nanocrystals on the structure and mechanical properties of poly
(lactic acid): a case of cellulose whisker-graft-polycaprolactone. Journal of
Applied Polymer Science 113, 3417–3425.
Lin, N., Huang, J., Chang, P.R., Feng, J.Yu., 2011. Surface acetylation of cellulose
nanocrystals and its reinforcing function in poly(lactic acid). Carbohydrate
Polymers 83, 23–88.
Frone, A.N., Berlioz, S., Chailan, J.F., Panaitescu, D.M., Donescu, D., 2011. Cellulose
fiber-reinforced polylactic acid. Polymer Composites 32, 976–985.
Huda, M.S., Drzal, L.T., Misra, M., Mohanty, A.K., 2006. Wood-fiber-reinforced poly
(lactic acid) composites: evaluation of the physicomechanical and morpholo-
gical properties. Journal of Applied Polymer Science 102, 4856–4869.
Jonoobi, M., Harun, J., Mathew, A., Oksman, K., 2010. Mechanical properties of
cellulose nanofiber (CNF) reinforced polylactic acid (PLA) prepared by twin
screw extrusion. Composite Science and Technology 69, 1742–1747.
Peterson, L., Kvien, I., Oksman, K., 2007. Structure and thermal properties of poly
(lactic acid)/cellulose whiskers nanocomposite materials. Composite Science
and Technology 67, 2535–2544.
662 T. Mukherjee et al. / Chemical Engineering Science 101 (2013) 655–662
Pei, A., Zhou, Q, Berglund, L.A., 2010. Functionalized cellulose nanocrystals as
biobased nucleation agents in poly(l-lactide) (PLLA)—crystallization and
mechanical property effects. Composites Science and Technology 70, 815–821.
Braun, B., Dogan, J.R., 2009. Single-step method for the isolation and surface
functionalization of cellulosic nanowhiskers. Biomacromolecules 10, 334–341.
Dubief, D., Samain, E., Dufrense, A., 1999. Polysaccharide microcrystals reinforced
amorphous poly(β-hydroxyoctanoate) nanocomposite materials. Macromole-
cules 32, 5765–5771.
Oksman, K., Mathew, A.P., Bondeson, D., Kvien, I., 2006. Manufacturing process of
cellulose whiskers/polylactic acid nanocomposites. Composite Science and
Technology 66, 2766–2784.
Pracella, M., Haque, M.M.U., Alvarez, V., 2010. Functionalization, compatibilization
and properties of polyolefin composites with natural fibers. Polymers 2,
554–574.
T. Mukherjee, N. Kao, R. K. Gupta, N.Quazi, S.N. Bhattacharya, Reinforcing Function
of Surface Acetylated Cellulose on Polylactic Acid (PLA) Based Biopolymer, 36th
Annual Condensed Matter and Materials Meeting, Wagga, 2012.
Mathew, A.P., Oksman, K., Sain, M., 2006. The effect of morphology and chemical
characteristics of cellulose reinforcements on the crystallinity of polylactic acid.
Journal of Applied Polymer Science 101, 300–310.
Das, K., Ray, D., Bandyopadhyay, N.R., Sengupta, S., 2010. Study of the properties of
microcrystalline cellulose particles from different renewable resources by XRD,
FTIR, nanoindentation, TGA and SEM. Journal of Polymers and the Environment
18, 355–363.
Furuhashi, Y., Yoshie, N., 2012. Stereocomplexation of solvent-cast poly(lactic acid)
by addition of non-solvents. Polymer International 60, 301–306.
Chen, X., Kalish, J., Hu, S.L., 2011. Structure evolution of α′-phase poly(lactic acid) by
addition of non-solvents. Journal of Polymer Science 49, 1446–1454.
Reinsch, V.E., Kelley, S.S., 1997. Crystallization of poly(hydroxybutrate-co-hydro-
xyvalerate) in wood fiber-reinforced composites. Journal of Applied Polymer
Science 64, 1785–1796.
Zhao, J., Morgan, A.B., Harris, J.D., 2005. Rheological characterization of polystyrene-
clay nanocomposites to compare the degree of exfoliation and dispersion. Polymer
46, 8641–8660.
Ray, S.S., Okamoto, M., 2003. Polymer/layered silicate nanocomposites: a review
from preparation to processing. Progress in Polymer Science 28, 1539–1641.
Dan, C.H., Lee, M.H., Kim, Y.D., Min, B.H., Kim, J.H., 2006. Effect of clay modifiers on
the morphology and physical properties of thermoplastic polyurethane/clay
nanocomposites. Polymer 47, 6718–6730.
Verney, V., Michel, A., 1989. Representation of the rheological properties of polymer
melts in terms of complex fluidity. Rheologica Acta 28, 54–60.
Cox, W.P., Merz, E.H., 1958. Correlation of dynamic and steady flow viscosities.
Journal of Polymer Science 28, 619–622.
Hatzikiriakos, S.G., Rathod, N., Muliawan, E.B., 2005. The effect of nanoclays on the
processibility of polyolefins. Polymer Engineering and Science 45, 1098–1107.